166 



BELL SYSTEM TECHNICAL JOURNAL 



To obtain reliable measurements throughout the length of the test 

 tube it was necessary to reduce standing waves to a negligible magni- 

 tude. This was accomplished by laying a strip of felt in the last 761 cm 

 of the tube, terminating the end with an acoustic resistance approxi- 

 mately equal to the characteristic impedance of the tube, and carefully 

 sealing up all joints along the tube. The pressure variation in the 

 standing wave was db 0.3 db, which corresponds to a reflection coeffi- 

 cient of 0.035. 



Fig. 2 — Apparatus for measuring extraneous frequencies generated in air carrying 

 intense sound waves. R, resistance substitute for receiver. 



The oscillator current was supplied to the loud speaker through a 

 low-pass filter and the measured harmonic content was found to be 

 73 db below the fundamental. Pressure measurements in the tube 

 close to the loud speaker indicated that the harmonics generated in the 

 measuring circuit and loud speaker were more than 50 db below the 

 fundamental pressure at 2000 bars. 



A calibration of the search transmitter was obtained by comparison 

 with a small condenser microphone whose diaphragm was exposed 

 directly to the sound wave at the same position on the test tube, see 

 Fig. 2. The calibrating microphone {C.T., Fig. 2) had been previously 



